Project Summary/Abstract Diagnosis and treatment monitoring of brain vascular diseases, such as stroke, arterial stenosis, vascular dementia, and small vessel diseases, requires accurate assessment of cerebrovascular parameters. There is substantial evidence that cerebrovascular reactivity (CVR), an index of vascular reserve reflecting the ability of the blood vessels to dilate or constrict, is one of the most sensitive markers for early detection of ischemia. In current clinical settings, CVR measurement requires a vascular challenge during imaging, involving either the injection of pharmacological agent (e.g. acetazolamide), inhalation of CO2 gas, or breath-hold. However, these strategies are not ideal in terms of both patient burden and costs, due to the complexity of the procedure, requirement of special equipment, and possible side-effect of physiological maneuver to induce vascular challenge. Therefore, development of a CVR technique that does not need an explicit vascular challenge will have a significant impact in cerebrovascular diseases. Resting-state BOLD MRI signal manifests spontaneous fluctuations with time. While portions of this signal have been exploited to map functional connectivity of brain networks and have revolutionized the fMRI field, another component of the signal fluctuation is of a vascular origin, due to spontaneous variations in breathing rate and depth that result in time-dependent changes in arterial CO2. This CO2 variation results in a fluctuation in BOLD signal, the amplitude of which is related to CVR. Our preliminary studies have strongly suggested the feasibility of mapping CVR using this CO2 fluctuation, as well as its ability to delineate vascular deficits in patients with intracranial arterial stenosis. Therefore, the central goal of the present study is to expand these preliminary findings and to develop and validate a CVR mapping technique based on MRI data acquired under resting-state, i.e. without gas challenges. First, we will develop acquisition and analysis methods to isolate the CO2 fluctuation component from resting-state BOLD signal for CVR mapping (Study 1 of Aim 1). Potential approaches to augment the sensitivity of the technique will be evaluated (Study 2 of Aim 1). Second, we will conduct simultaneous MRI, electroencephalogram (EEG), and end-tidal CO2 recordings to verify that the signal measured with our method contains minimal neural contribution but predominantly CO2 fluctuation, and validate the proposed CVR mapping technique with the gold-standard CO2-inhalation method (in Aim 2). Finally, we will conduct initial clinical demonstration of this technique in patients with intracranial arterial stenosis (in Aim 3). Impact: The impact on clinical practice is that patients with cerebrovascular diseases, including ones whose are unable to comply with vascular challenges, can receive CVR imaging during a standard MRI session. This will greatly enhance the utility of CVR in clinical diagnosis and prognosis.